Etched develops artificial intelligence hardware and software systems for inference workloads. The company designs specialized AI chips alongside supporting components, including racks, interconnects, software, and manufacturing processes, to optimize the deployment of large-scale AI models. Its technology focuses on improving throughput, latency, power efficiency, and memory performance for transformer-based and mixture-of-experts inference workloads. Etched also develops rack-scale inference systems that integrate compute, memory, networking, and cooling technologies into a unified platform. The company works on the validation, production, and deployment of its inference infrastructure for customers running large language models and other advanced AI applications.